Build it with AI
Build one view of spend, traffic, leads, conversion, and revenue influence.
Create a marketing dashboard around the channels and KPIs your team actually manages.
Introduction
What a marketing dashboard has to hold.
Most teams end up with a marketing dashboard the same way: a monthly spreadsheet assembled from four ad platforms and an analytics export. Reading each platform’s own dashboard works while there are two channels. At six, the monthly report is a manual assembly job, and every platform is grading its own homework using a different definition of a conversion.
The assembly takes two days, so by the time anyone reads it the month it describes is already over. There is no shared definition of a conversion across platforms, no place where spend and pipeline meet, and no record of what the numbers were before a platform retroactively restated them.
What follows covers building a marketing dashboard: the records it holds (spend, impressions, traffic, leads, conversion, and influenced pipeline by channel), the systems it reads (GA4 and HubSpot), and what it does not fix.
The problem
Six channel dashboards and no view of the whole.
Each ad platform reports on its own contribution using its own attribution window, which is why the sum of platform-reported conversions reliably exceeds the number of actual customers. Web analytics is neutral and cannot see spend.
The records are spend, impressions, traffic, leads, conversion, and influenced pipeline by channel, and the authoritative copy of most of them already lives in GA4 or HubSpot. Platform-reported conversions total 340, the CRM shows 190 new opportunities, and the monthly report reconciles them by picking whichever number supports the point being made.
The cost is not the inconvenience: budget is renewed on channels that produce leads and no pipeline.
You're likely here because
- The channel numbers add up to more conversions than the business had
- The assembly takes two days, so by the time anyone reads it the month it describes is already over.
- When it is wrong, budget is renewed on channels that produce leads and no pipeline
What gets built
Launch builds it, Grow operates it.
Built in Launch
- • Channel KPIs
- • Campaign views
- • Conversion reporting
Operated through Grow
- • Lead context
- • Attribution
- • Pipeline linkage
Systems it reads
- • GA4
- • HubSpot
- • Google Sheets
The record model
What the marketing view has to reconcile.
- Business-defined conversion
- One definition applied to every channel, stated on the view. Platform-defined conversions are not comparable to each other and reliably sum to more than the business had.
- Spend, platform-reported and invoiced
- Both, because they diverge more often than teams assume and the reconciliation is monthly rather than continuous.
- Attribution model and window
- Named on every view. Changing either changes which channel wins, so a figure without its model is an argument waiting to happen.
- Unattributed conversions
- As their own category. Distributing them proportionally is the standard practice and it systematically flatters whichever channel is easiest to track.
- Snapshot date
- Because platforms restate history. Without a snapshot policy, last month’s report will not reproduce next month and nobody will be able to say why.
- Channel and campaign taxonomy
- Owned and versioned rather than inherited from whatever each platform calls things, so a comparison across channels is comparing the same shapes.
- Join rate to the CRM
- The share of conversions that reached a CRM record. It is the honest ceiling on any claim about pipeline influence.
How it runs
From platform reports to one channel-agnostic view.
Step 01
Describe what a marketing dashboard has to do
Define a conversion once, in business terms, and hold every channel to it. The definition is the dashboard; the charts are a rendering of it.
Step 02
Connect the systems of record
Analytics supplies sessions and on-site behaviour, the CRM supplies what became pipeline, and the spend figures come from the platforms. The join is the work, and it should be visible.
Step 03
Build the operating surface
Channel KPIs against one conversion definition, campaign-level views, and the path from spend through to pipeline influence — with the attribution model named on the view.
Step 04
Start narrow
Spend and business-defined conversions for the two largest channels, on one screen, with the unmatched share shown. Two channels done honestly beats six done by summing platform reports.
Step 05
Route the exceptions
A channel whose cost per business-defined conversion moves beyond its normal band surfaces for review rather than being noticed in the monthly cycle.
Step 06
Measure cost per opportunity by channel, not cost per lead
Cost per business-defined conversion by channel, with the unmatched rate alongside. Platform-reported cost per conversion is a different and consistently more flattering number.
Implementation path
Building a marketing view that survives a channel change.
- 01
Write the conversion definition first and check it against the CRM. If marketing’s conversion is a form fill and the business cares about qualified opportunities, every figure downstream will be arguing about a different thing.
- 02
Baseline the discrepancy: platform-reported conversions versus business-defined ones for last quarter. That gap is the reason the dashboard is worth building and the number to quote when justifying it.
- 03
Connect analytics and the CRM before pulling in spend. Spend is the easy part and adding it first produces a dashboard that looks complete while resting on an unverified join.
- 04
Publish with the attribution model named and the unmatched share visible, from the first version. Retrofitting honesty into a dashboard people already trust is much harder than building it in.
- 05
Build the narrowest useful version first: spend against pipeline created for the two channels that carry most of the budget.
- 06
Agreeing the conversion definition and checking it against the CRM takes a morning and is where the disagreement surfaces. Connecting analytics and the CRM is the first week; spend is easy and should come last so the join is verified before the numbers look complete. Two channels done honestly within a fortnight beats six done by summing platform reports.
- 07
Once cost per business-defined conversion is live for the two largest channels, add cohorted payback rather than period ratios — it stops a long sales cycle from making recent spend look free. Creative-level performance comes after, and only where volume supports it.
Controls
Controls that matter.
Control 01
One conversion definition applied across every channel, stated on the dashboard rather than held in someone’s head
Control 02
The attribution model and its window named on each view, since changing either changes which channel wins
Control 03
Spend figures reconciled against invoices at least monthly, because platform-reported spend and billed spend diverge more often than teams assume
Examples
Three arguments that get shorter.
The channel that looked best in its own dashboard
Holding every channel to one business-defined conversion frequently reverses the ranking, because the platform reporting most generously is not the one contributing most.
The monthly report assembly
The two days of copying between platform exports disappear. What remains is deciding what the numbers mean, which was the part with any judgement in it.
The budget conversation
Cost per business-defined conversion by channel, with the unmatched rate shown, turns a debate between two advocates into a comparison with stated uncertainty.
How it goes wrong
Three ways channel reporting misleads.
Platform-reported conversions are summed across channels and the total exceeds the number of actual customers.
Hold every channel to one business-defined conversion. The sum being wrong is not a rounding issue — each platform counts every conversion it can claim, and the overlap is the whole discrepancy.
A dashboard built on a platform API cannot reproduce a figure quoted in last quarter’s board pack.
Snapshot the figures you report on. Platforms restate historical data routinely, and a board number that changes retroactively costs more credibility than it is worth to avoid the storage.
Unattributed conversions are spread across the paid channels so the numbers add up neatly.
Show them separately. Brand and word of mouth are real acquisition and are undercounted by every channel-attributed view; hiding that makes the measurable channels look better than they are.
Limitations and considerations
What channel data cannot honestly tell you.
- Privacy changes, consent rejection, and cross-device journeys mean a meaningful share of conversions cannot be attributed to a channel at all. That share belongs on the dashboard rather than being distributed proportionally into the channels that can be measured.
- Brand and organic effects show up as unattributed conversions and are systematically undercounted by any last-touch view. A dashboard cannot fix this; it can avoid pretending otherwise.
- Platforms restate historical figures. Anything built on a platform API needs a snapshot policy, or last month’s report will not reproduce next month.
- With one or two channels, the platform dashboards plus a monthly spreadsheet are proportionate. If the conversion definition cannot be agreed, build that agreement first — the dashboard will otherwise become the venue for the argument rather than its resolution.
- Connector coverage varies: GA4, HubSpot, Google Sheets are representative rather than guaranteed, and the fields exposed depend on your workspace permissions.
FAQ
Build a marketing dashboard with AI: common questions.
Why do our numbers never match the ad platforms?
Because each platform attributes using its own window and its own definition, and counts a conversion it can claim. Holding every channel to one business-defined conversion is what makes the numbers comparable — and it reliably produces a smaller, more useful total.
Do we need a data warehouse?
Not for a first version reading analytics, the CRM, and platform spend directly. A warehouse earns its place when you need history that survives platform restatements, which is a real problem but a later one.
How should we handle unattributed conversions?
Show them as their own category. Distributing them proportionally across measurable channels is the standard practice and it systematically flatters whichever channel is easiest to track.
What if marketing and sales define a lead differently?
Then the dashboard will make that visible in its first week, which is the most useful thing it does. Resolve the definition before publishing widely, or the tool becomes the venue for the argument rather than the resolution of it.
What should the first version contain?
Spend against pipeline created for the two channels that carry most of the budget. Everything else waits until that one is genuinely used.
How will we know whether it worked?
Measure cost per opportunity by channel, not cost per lead against the baseline taken before anything changed.
Start with ARIA
Ask ARIA to build it.
Describe the website, application, workflow, or operating surface you need. ARIA plans, connects, builds, tests, and keeps refining it — inside the permissions you set.
- ARIA acts only through the systems and permissions you connect.
- Connections use scoped credentials you can change or revoke.
- Actions are recorded, and consequential ones can require approval.
Start here
Build a marketing dashboard around the process you actually run.
Define a conversion once in business terms, hold every channel to it, and put the unattributed share on the chart.